Destruction, Not Addition

The intro left you with a slogan: growth is replacement, not accumulation. A slogan is easy to nod along to and easy to forget the moment a beloved company dies and you feel surprised anyway. So this lesson does something the slogan can’t — it makes the claim into a picture you can watch move. There is exactly one shape at the heart of creative destruction, and once you can see it, you’ll spot it everywhere: two S-curves, one rising and one falling, crossing like an X. That crossing is the whole model.

The two curves are one event

Start with the picture almost everyone carries and doesn’t know they’re carrying. When we imagine a new technology “taking off,” we picture a single line climbing upward — adoption going up and to the right. Clean, optimistic, and incomplete. It leaves out the other half of the frame: the line for the thing being replaced, sliding down at the same time.

Draw both and the shape snaps into focus. The challenger’s adoption traces a rising S-curve. The incumbent’s installed base traces a falling one. And here’s the part that matters: they are mirror images. At every moment, the share the challenger has gained is almost exactly the share the incumbent has lost. Add the two together and you get roughly a constant — the whole market — because the market didn’t grow a second copy of itself. The same customers just walked from one seller to the other.

Why must they mirror? Because the resources are conserved. A household watches only so many films a month; an hour spent streaming is an hour not spent at the rental store. A dollar of a customer’s budget spent on the new thing is a dollar not spent on the old. The skilled worker retrained onto the new line is a worker no longer on the old one. The challenger doesn’t conjure its customers, revenue, and labour out of thin air — it takes them, one crossover at a time, from the incumbent that had them. That is why Schumpeter insisted creation and destruction are not two events that happen to coincide. They are one event seen from two sides.

Now drive it. Below is the wave itself. The green curve is the challenger climbing; the red curve is the incumbent’s installed base collapsing beneath it. Move the dials and watch how the timing changes while the shape never does.

Three things to notice as you play with it:

• The crossover point at 50%. The moment the challenger passes half the market, the black dot lights up. This is the point of no return. Past it, the market has decided: the challenger now has the larger installed base, the bigger revenue, the better unit economics from scale, and the momentum. The incumbent isn’t merely losing — it is effectively doomed, even though it may still be a large, cash-generating business for years. Doom in this model isn’t bankruptcy tomorrow; it’s the point after which no amount of running the old playbook can win the market back.
• The dials change the clock, not the outcome. Crank the challenger’s improvement rate and the crossover slides earlier; raise the incumbent’s moat — switching costs, brand, scale — and it slides later. But you’re moving when, not whether. Only an enormous moat against a slow challenger keeps the crossover off the chart entirely (more on that limit at the end).
• The dashed “next wave.” See the faint dashed curve rising near the right edge? That is the innovation that will one day destroy the challenger in turn — funded by the very resources this crossover just freed up. Creative destruction doesn’t end at a winner. It just reloads.

Why it must be destruction, not addition

Hold up the two mental pictures from the intro side by side, because this section is where the choice between them actually bites.

In the warehouse picture, the new technology is a fresh shelf added to the stockroom. The old shelves stay exactly where they were; growth is just the pile getting taller. Nothing is thrown out. In the forest picture, the new growth is a seedling that becomes a giant only by taking the light, water, and soil the old tree was using. It grows by shading out what stood there before.

The forest picture isn’t just prettier — it’s the correct one, and here is the airtight reason. A new product has to be bought, with real money, by real customers who have finite budgets. Its factory runs on capital that had to come from somewhere. Its workers have skills that had to be trained and hours that can only be spent once. None of that is free-floating. Every unit of it was, moments ago, attached to the incumbent. The challenger can’t rise on empty air; it rises on the incumbent’s resources. So the incumbent’s decline isn’t an unlucky side effect of the challenger’s success — it is where the success comes from. Delete the destruction and you’ve deleted the fuel.

And “destruction” here is not a genteel euphemism for “moved upward.” This is the part comfortable retellings skip. When film photography collapsed, the film chemist’s specialised expertise did not get promoted to a better job — it became obsolete. The projectionist threading reels at the multiplex, the craftsman stitching horse tack when the automobile arrived, the typesetter arranging metal type before desktop publishing — these people didn’t ascend a ladder. Their specific skills, and often their firms and their towns, were genuinely destroyed. New and different jobs appeared elsewhere, for different people with different skills, which is real and matters — but it is not the same as the displaced individual being carried gently upward. Conflating “the economy created new jobs” with “the destroyed worker was fine” is the single most common way this model gets sanitised into a fairy tale.

A worked timeline: streaming vs. video rental

Abstractions are slippery; let’s watch one real crossover happen slowly enough to see every stage. Take home video — specifically Blockbuster (the incumbent) versus Netflix (the challenger).

At its peak around 2004, Blockbuster ran roughly 9,000 stores and was a fixture of the weekend: you drove over, browsed the shelves, rented a tape or DVD, and — crucially for its economics — paid a late fee if you returned it a day late. Those late fees weren’t a nuisance line item; they were a substantial slice of the profit. That’s the incumbent’s business at full health: physical stores, physical inventory, and a revenue model with a sharp edge customers quietly resented.

Netflix began climbing the bottom of its S-curve as DVD-by-mail: no store, no due date, no late fee — a red envelope in your mailbox. Early on it looked tiny and worse in obvious ways (you waited a day or two for the disc). Then the curve steepened as Netflix pivoted to streaming: the movie arrived instantly, for a flat monthly price, on the screen already in the room. Now walk the crossover. The households doing the crossing are the same households — the same people who wanted to watch a film on Friday night. The demand never changed. Only the delivery curve did. Every family that switched to streaming was a family that stopped driving to the store, which meant a shelf of DVDs going unrented and a late fee never collected. The challenger’s rise was the incumbent’s decline, dollar for dollar and household for household.

Here is the crossover as a table — the same market, read at three moments:

Blockbuster filed for bankruptcy in 2010. Notice what did and didn’t change. The underlying want — watch a film at home — was rock steady the whole way through; nobody started wanting fewer movies. What got destroyed was one specific way of delivering that want, along with the 9,000 storefronts, the shelving, the late-fee revenue, and the jobs attached to them. That is the crossover in the wild: same customers, same demand, one curve replacing another beneath them.

The S-curve shape itself

We keep saying “S-curve.” Why that shape — slow, then explosive, then flat — and not a straight ramp? The three phases each have a plain economic reason, and understanding them tells you how steep any given wave will be.

The slow start (the bottom of the S). Early on, the new technology is usually worse, more expensive, or both. Netflix-by-mail made you wait for a disc; the first digital cameras were pricey and grainy; early cars were unreliable and there were no petrol stations. Only a few tolerant early adopters bother, so adoption crawls. The incumbent barely notices — which is exactly why it feels safe.

The explosive middle (the steep part). Then the challenger crosses a threshold: good enough, and cheaper. The moment it clears the bar for the mainstream buyer while undercutting the incumbent on price or convenience, ordinary supply and demand takes over and adoption goes vertical. Each convert makes switching more normal for the next (word of mouth, better infrastructure, falling prices from scale). This is where the crossover happens and where the incumbent’s revenue falls off a cliff.

The saturation at the top (the flattening). Eventually almost everyone who’s going to switch has switched. There’s no one left to convert, so the curve levels off near the top of the market. The challenger is now the incumbent — and the faint dashed “next wave” on the chart is already stirring below it.

The steepness of that middle section is set by how fast the challenger improves and how hard it undercuts. A challenger that gets dramatically better and cheaper every year (a big improvement-rate dial) produces a near-vertical crossover — a rout. A challenger that inches forward against a strong incumbent produces a gentle, drawn-out slope. Same S, different tilt. Watch it directly here: this second wave is a fast challenger against a paper-thin moat — the shape a genuine rout takes.

What the crossover does NOT promise

Now the honest limits, because a model you can’t break is a model you’ll misuse. The crossover is a strong tendency, not a stopwatch and not an iron law. Two cautions.

First: crossing over does not make the challenger safe forever. Passing 50% wins this wave — it does not buy immunity from the next one. Remember the dashed curve: the resources this crossover just freed are already funding whatever will one day do to the challenger exactly what it did to the incumbent. Netflix beat Blockbuster and then found itself in a knife fight with a dozen other streaming services. Winning is a lease, not a deed. (We’ll make this rhythm explicit later, under the Red Queen — no moat is permanent, and standing still is falling behind.)

Second: some incumbents genuinely hold — for a long time. A slow challenger against a strong, real moat may never cross 50% inside any horizon a human planner cares about. That’s the “fortress” setting on the dial, and it isn’t cheating — switching costs, network effects, regulation, and entrenched standards are real forces that have kept incumbents on top for decades. The model predicts a direction, not a date. When you see a wave, you know which way it points; you do not automatically know it will arrive on schedule. (The incumbents who actually adapt, and the moats that genuinely hold, get their own lesson at the end of the course.)

Put the two cautions together and you get the mature reading of the model: creative destruction tells you the shape and the direction of change with unusual reliability, while leaving the speed genuinely open. Use it to see who’s on the falling curve and why — not to bet on precisely which year the crossover lands.

Let’s pull the whole lesson into one map before moving on.

Where this goes next

You can now see the central claim instead of just reciting it: growth is the market crossing from a falling curve to a rising one, the same resources changing hands, the destruction and the creation welded into a single event at the crossover. That picture is the spine of everything that follows — every case in this course is a wave you can sketch.

But we’ve been describing the wave as if it just happens. It doesn’t. Something has to light the match — someone has to build the challenger, price it under the incumbent, and drag the mainstream across. The next lesson opens up the engine of the gale: the entrepreneur, and Schumpeter’s five kinds of innovation (new goods, new methods, new markets, new sources of supply, and new ways of organising). Innovation, crucially, is not the same thing as invention — and knowing the difference is what lets you tell which challengers will actually climb the curve and which will stall at the bottom of the S. The wave showed you what happens. Next we find out what makes it move.